Projects

Rohingya refugee camp management system
  • Managed 500k+ OracleDB tables, developed web apps with JavaScript, PHP, SQL
  • For server security used UFW and Nginx for optimization, Certbot for Secure SSL
  • Admin and user-level access, the director can give user-level access to other officers
JavaScript
PHP
SQL
OracleDB
Nginx
Certbot
VoIP custom CID calling system
  • Secured client-server connection with Asterisk PBX on Ubuntu
  • Automated PBX server monitoring and alerts with Bash script
  • Developed the system using Raspberry Pi 4 and optimized network infrastructure
Raspberry Pi 4
Asterisk PBX
Ubuntu
Bash Script
An Unique Research on the Application of OpenStack to Build up a FHRVC for Representing a Network Security Technology Platform
RecomBook
A book recommender system using user-based Collaborative Filtering Algorithm.
  • The dataset contains 278,858 users, 1,149,780 ratings and 271,360 books
  • Content-based filtering Algorithm is used to provide popular 50 books to users
Python
Flask
Numpy
Pandas
JavaScript
Bootstrap
SocialAppREST
A social platform REST-API with a voting system, JWT authentication and authorization. The complete software life cycle of this project has been implemented.
  • Created user login, registration, and authentication with JWT token
  • Used PostgreSQL database and secured API endpoints with user authentication
  • SQLAlchemy and Alembic to perform CRUD operations based on Pydantic models
  • Unit testing with PyTest, API testing with Postman, and App dockerized with Dockerfile
  • CI/CD pipeline with GitHub Actions workflow for automated deployment
  • Deployed on Heroku and Production server, tweaked Nginx for optimal performance
Python
FastAPI
Pydantic
SQLAlchemy
JWT OAuth 2.0
PostgreSQL
Alembic
PredBrainTumor
A deep learning model for predicting brain tumor from MRI images using TensorFlow Convolutional Neural Network (CNN). Transfer learning is used to train the model. The model has four classes: meningioma, glioma, pituitary tumor, and no tumor with 98% prediction accuracy.
  • Data Scaling and Normalization using Keras
  • Data prepared using Numpy and OpenCV into 150 x 150
  • EfficientNetB0 model, to use weights from ImageNet dataset
  • TensorBoard, ModelCheckpoint and ReduceLROnPlateau used for callback functions
Python
TensorFlow
ImageNet
Scikit-learn
Keras
Matplotlib
Numpy
OpenCV
Flask
2dGeoTrans
PhD Task: Geometrically transform the stereoscopic 2D image (object position) to a new position based on the photographic composition rules.
  • OpenCV and Numpy for image processing
  • combination of ORB and Viola-Jones detector framework for feature matching
  • Rule of thirds implemented with Haarcascade classifier as Hybrid Detector

Contact: Y A Joarder
Python
OpenCV
Numpy
Juptyer Notebook
JavaScript
Haarcascade Algorithm
CS Study Guide
  • Presenting globally standardized self-taught study plans on GitHub covering Computer Science, Data Science, Machine Learning, Web Development, Artificial Intelligence, Higher Education, and Data Structures and Algorithms. These meticulously curated resources offer comprehensive pathways for self-learners worldwide, empowering individuals to excel in the dynamic fields of technology and innovation.
Computer-science
Data-science
Machine-learning
Web-development
Artifical-intelligence
Higher-education
Data-structures-and-algorithms